Abstract
Introduction
Atrial fibrillation (AF) is the most common cardiac arrhythmia in clinical settings known to impair cardiac function. Genome‐wide association studies identified SNPs on chromosome 4q25 to be associated with AF. Till date no information is available from India related to the association of these variants with AF.
Materials and Methods
Two hundred and sixty‐seven individuals comprising of 170 patients with Coronary Artery Disease (CAD), 41 patients with AF, and 56 healthy controls were genotyped for rs10033464 and rs2200733 at 4q25 locus.
Results
Strong association of rs10033464 with AF was observed on comparison with control groups (OR: 2.59; 95% CI: 1.08–6.21; P: 0.031) and between post‐coronary artery bypass grafting (CABG) AF and control with dominant genetic model (OR: 4.73; 95% CI: 1.50–14.89; P: 0.0071). Comparison of post‐CABG AF with CAD also indicated association (OR: 2.73; 95% CI: 0.9–7.56; P value: 0.05). In contrast, the rs2200733 C>T variant did not show any association with post‐CABG AF, but with lone AF with the ‘T’ allele being associated with increased risk was seen (OR: 2.80; 95% CI: 1.08–7.24; P value: 0.042).
Conclusion
In conclusion, the rs10033464 (T) allele is associated with the risk of post‐CABG AF and the rs2200733 (T) with lone AF.
Keywords: 4q25, association study, atrial fibrillation, coronary artery disease, single nucleotide polymorphisms
Background
Atrial fibrillation (AF) is the most common arrhythmia in the clinical setting and an independent risk factor for stroke. In AF, the normal regular electrical impulses generated by the sinoatrial (SA) node are overwhelmed by disorganized electrical impulses usually originating in the roots of the pulmonary veins, leading to irregular conduction of impulses to the ventricles which generate the heartbeat 1, 2. From the epidemiological perspective, AF represents the most serious problem. In a retrospective study on a heterogeneous racial mix of US medicare beneficiaries, an incidence between 27 and 28.3 per 1,000 person years has been consistently reported with little change over the past two decades 3.
In a vast majority of patients AF may be caused due to underlying cardiac or non‐cardiac conditions; however, a subset of patients with AF on evaluation does not reveal any abnormalities. These patients are considered to suffer from ‘lone’ AF, i.e., AF without an underlying (cardiovascular) disorder. The risk factors for AF are multi‐factorial and include male gender, advancing age, hypertension, diabetes, obesity, coronary artery disease (CAD), renal failure, chronic obstructive pulmonary disease, congestive heart failure, and valvular heart disease 4.
Genetic variants have also been associated and the first genome‐wide association study (GWAS) identified association between genetic variants on 4q25 locus and AF. In particular, two single‐nucleotide polymorphisms (SNPs), rs2200733 and rs10033464, have been pointed out as markers most strongly associated with the disease 5. The closest gene is PITX2, which encodes a protein that is critical in the development of the heart for determining left–right asymmetry, SA node formation, and the differentiation of the left atrium.
The association between AF and rs2200733/rs10033464 has been observed in Icelandic, Swedish, USA, and Chinese populations, as well as the population of the Framingham Heart Study (FHS) and several other Western European populations 5, 6, 7, 8, 9, 10, 11. In addition, these polymorphisms are also associated with postoperative AF (PoAF), postoperative strokes, long‐term development of AF, long‐term stroke incidence, and long‐term survival after coronary artery bypass grafting (CABG) 6, 7, 8, 9, 10. These variants could stratify postoperative, long‐term AF risk in these patients, and potentially affect the choice of preoperative therapy used to decrease the risk of PoAF.
However, many independent replication studies are essential to unequivocally validate this association in other populations worldwide. Till date no information is available from India and this study aims to address this lacunae.
Materials and Methods
Subject Groups
The study cohort comprised of 267 individuals of indo‐aryan ethnicity, which included 170 patients with CAD, 41 patients with AF, and 56 individuals who were deemed normal.
The AF status of each individual was confirmed by Resting/stress Electrocardiogram testing and/or Treadmill Test. Individuals with history of Rheumatic Heart Disease, or show heart valve stenoses like mitral valve stenosis, or have undergone valvular surgery were excluded. EDTA‐Blood specimens were obtained from each individual. An informed consent and detailed case record form pertaining to information on demographics, medical history, and risk factors such as diabetes mellitus, hypertension, events of cardiac stroke, myocardial infarctions, lipid profiles, family history, etc., was obtained from each of the participants through perusal of their medical records.
DNA Extraction
Genomic DNA was extracted from the collected EDTA whole blood samples using QIAamp (R) DNA extraction kit following the manufacturer's protocol (Qiagen, Hilden, Germany).
Genotyping
Genotyping was performed for the rs10033464 and rs2200733 by direct sequencing of the products obtained after amplification using the primers described below. The PCR cycling conditions were as follows: initial denaturation at 95°C for 5 min; 35 cycles of dentauration at 95°C for 30 sec, annealing at 60°C for 30 sec, and extension at 72°C for 30 sec; and final extension at 72°C for 10 min. Sequencing was done bidirectional on the Automated ABI prism 3100 Avant Genetic Analyzer (Applied Biosystems Inc., Foster city, CA) using ABI prism BigDye terminator kit (version 3.1).
rs10033464 F 5′‐3′: AGTTCTTGAGGAAAAGCCTGTTG
rs10033464 R 5′‐3′: TCACAGCCGAGGATGTCTTTAGT
rs2200733 F 5′‐3′: TGGATGAATGAATGAACTAGTGC
rs2200733 R 5′‐3′: TCACACTTGACTCTCTACATGTCC
Statistical Analysis
Allele frequency was calculated as the number of occurrences of the test allele in the population divided by the total number of alleles. Any deviation in the genotype frequencies from the Hardy–Weinberg equilibrium was assessed by Fischer's exact test. Chi‐square tests were used for comparison of binary variables across groups. Routine statistical analysis was carried out with the SPSS v 15 software (SPSS Inc., Chicago, IL). SNPStat online software tool 12 was applied to determine the association of genotypes with the AF. The association between AF and each SNP was tested using logistic regression with a genotypic genetic model.
Results
General Characteristics of Cases and Control Groups
Table 1 displays means and standard deviations for the study subjects for relevant biochemical characteristics as well as risk factors. Statistically significant differences were seen in the presence of stroke/myocardial infarction in family across all the three groups and diabetes mellitus among AF patients as compared to CAD and controls group (P < 0.05).
Table 1.
General Characteristics of Cases and Controls
| General characteristics | AF group (n = 41) | CAD group (n = 170) | Control group (n = 57) |
|---|---|---|---|
| Age (in Years) | 69.7 ± 10.2 | 59.6 ± 9.7 | 49 ± 10.1 |
| Gender (male) | 91.2 | 86.7 | 73.2 |
| BMI (kg/m2) | 25.8 ± 3.5 | 26 ± 3.7 | 25.51 ± 3.9 |
| Smoking status (up to 10 cig/day) | 4 (10) | 18 (10.6) | 7 (12.2) |
| Alcohol (at least 120 ml/week) | 1 (0.2) | 21 (12.3) | – |
| Diabetes mellitus | 17 (41.5) | 91 (52.2)d | 5 (26.3)a |
| Hypertension | 25 (61) | 110 (64.7) | 12 (50)b |
| Atrial Fibrillation (AF) | |||
| (a) Lone AF | 24 (59) | – | – |
| (b) Post‐CABG AF | 17 (41) | ||
| Lipid profilesc | |||
| (a) Total cholesterol (gm/dl) | 161 ± 23.4 | 161.7 ± 45.4 | 191.5 ± 30.5 |
| (b) Triglycerides (gm/dl) | 118.3 ± 33 | 142 ± 74.3 | 134.5 ± 48.3 |
| (c) HDL‐Cholesterol (gm/dl) | 43.6 ± 10.3 | 43.1 ± 9.7 | 40 ± 7.4 |
| (d) LDL‐Cholesterol (gm/dl) | 101.3 ± 29 | 94.7 ± 39.4 | 123 ± 25.5 |
| Family history | |||
| (a) Stroke/MI | 15 (36.6)d | 47 (28)d | 8 (14)d |
| (b) Diabetes mellitus | 20 (49)d | 38 (22.3) | 12 (21) |
| (c) Hypertension | 27 (66) | 30 (17.6) | 8 (14) |
| (d) Other inherited disorders | – | 2 (12) | 9 (15.7) |
Data are represented as n (%) or mean ± S.D. AF indicates atrial fibrillation; BMI, body mass index; CABG, coronary artery bypass graft, HDL, high‐density lipoproteins; LDL, low‐density lipoproteins; MI, myocardial infarction; CAD, Coronary Artery Disease.
Status data for diabetes available for 19 controls.
Status data for hypertension available for 24 controls.
Data for lipid profiles available for 26 AF patients, 130 CAD patients, and 50 controls.
P value <0.05 when cases AF and CAD were compared to controls.
Genotypic and allelic frequencies are shown in Table 2, both the rs10033464 and rs2200733 were found to be in Hardy–Weinberg equilibrium using Fischer's Exact test.
Table 2.
Genotype and Allelic Frequencies in AF, CAD, and Control Groups
| SNP | Genotype frequencies n (%) | Major | Minor | HWE | ||
|---|---|---|---|---|---|---|
| rs10033464 | GG | GT | TT | G | T | P‐value |
| CAD (n = 170) | 112 (66) | 51 (30) | 7 (4) | 0.80 | 0.20 | 0.63 |
| AF (n = 41) | 23 (56) | 17 (41.5) | 1 (2.5) | 0.77 | 0.23 | 0.66 |
| Controls (n = 56) | 43 (77) | 13 (23) | – | 0.88 | 0.12 | 1 |
| rs2200733 | CC | CT | TT | C | T | |
|---|---|---|---|---|---|---|
| CAD (n = 166) | 141 (85) | 22 (13) | 3 (2) | 0.91 | 0.09 | 0.92 |
| AF (n = 41) | 32 (78) | 8 (20) | 1 (2) | 0.88 | 0.12 | 0.54 |
| Controls (n = 57) | 42 (73.6) | 15 (26.3) | – | 0.87 | 0.13 | 0.58 |
AF, Atrial fibrillation; CAD, Coronary Artery Disease.
Further to determine the association of the rs10033464 variant with AF we used the logistic regression analysis with different genetic models. The univariate model showed strong association of rs10033464 ‘T’ allele with AF risk when compared between AF and control groups (OR, 2.59; 95% CI, 1.08–6.21; P = 0.031), and between post‐CABG‐AF and control with the dominant model (OR, 4.73; 95% CI, 1.50–14.89, P = 0.0071). No association of rs10033464 was seen when AF patients were compared with CAD group (OR, 1.50; 95% CI, 0.75–03.00; P = 0.26). Comparison of post‐CABG AF patients and CAD group indicated borderline significance with the dominant model (OR, 2.73; 95%CI, 0.9–7.56, P = 0.05) (Table 3).
Table 3.
Univariate Logistic Regression Analysis to Determine Association of rs10033464G>T with Atrial Fibrillation
| rs10033464G>T | ||||||
|---|---|---|---|---|---|---|
| Genotype | Control (%) | AF (%) | OR (95% CI) | P‐value | AIC | |
| Dominant | G/G | 43 (76.8) | 23 (56.1) | 1.00 | 0.031a | 131.5 |
| G/T‐T/T | 13 (23.2) | 18 (43.9) | 2.59 (1.08–6.21) | |||
| Genotype | Control (%) | Post‐CABG‐AF (%) | OR (95% CI) | P‐value | AIC | |
| Dominant | G/G | 43 (76.8) | 7 (41.2) | 1.00 | 0.0071a | 76 |
| G/T‐T/T | 13 (23.2) | 10 (58.8) | 4.73 (1.50–14.89) | |||
| Genotype | CAD (%) | AF (%) | OR (95% CI) | P‐value | AIC | |
| Dominant | G/G | 111 (65.7) | 23 (56.1) | 1.00 | 0.26 | 210.1 |
| G/T‐T/T | 58 (34.3) | 18 (43.9) | 1.50 (0.75–3.00) | |||
| Genotype | CAD (%) | Post‐CABG AF (%) | OR (95% CI) | P‐value | AIC | |
| Dominant | G/G | 111 (65.7) | 7 (41.2) | 1.00 | 0.05a | 113.9 |
| G/T‐T/T | 58 (34.3) | 10 (58.8) | 2.73 (0.99–7.56) | |||
| Genotype | CAD (%) | Lone AF (%) | OR (95% CI) | P‐value | AIC | |
| Dominant | G/G | 110 (65.5) | 16 (66.7) | 1.00 | 0.91 | 148.7 |
| G/T‐T/T | 58 (34.5) | 8 (33.3) | 0.95 (0.38–2.35) | |||
OR, odds ratio; 95% CI, 95% Confidence Interval; AIC, Akaike Information Content; AF, Atrial fibrillation; CAD, Coronary Artery Disease.
Statistically significant at P ≤ 0.05.
In contrast, the rs2200733 C>T variant did not show an association with AF, or post‐CABG AF (Table 4) when compared to CAD, but an association with lone AF with the ‘T’ allele being associated with increased risk (OR, 2.80; 95% CI, 1.08–7.24; P, 0.042) (Table 4).
Table 4.
Univariate Logistic Regression Analysis to Determine Association of rs2200733 C>T with Atrial Fibrillation
| rs2200733 C>T | ||||||
|---|---|---|---|---|---|---|
| Genotype | Control (%) | AF (%) | OR (95% CI) | P‐value | AIC | |
| Dominant | C/C | 42 (73.7) | 31 (75.6) | 1.00 | 0.83 | 137.2 |
| C/T‐T/T | 15 (26.3) | 10 (24.4) | 0.90 (0.36–2.28) | |||
| Genotype | Control (%) | Post‐CABG‐AF (%) | OR (95% CI) | P‐value | AIC | |
| Dominant | C/C | 42 (73.7) | 14 (82.3) | 1.00 | 0.45 | 83.2 |
| C/T‐T/T | 15 (26.3) | 3 (17.6) | 0.60 (0.15–2.38) | |||
| Genotype | CAD (%) | AF (%) | OR (95% CI) | P‐value | AIC | |
| Dominant | C/C | 141 (84.9) | 31 (75.6) | 1.00 | 0.17 | 208.2 |
| C/T‐T/T | 25 (15.1) | 10 (24.4) | 1.82 (0.79–4.17) | |||
| Genotype | CAD (%) | Post‐CABG AF (%) | OR (95% CI) | P‐value | AIC | |
| Dominant | C/C | 141 (84.9) | 14 (82.3) | 1.00 | 0.78 | 117.1 |
| C/T‐T/T | 25 (15.1) | 3 (17.6) | 1.21 (0.32–4.51) | |||
| Genotype | CAD (%) | Lone AF (%) | OR (95% CI) | P‐value | AIC | |
| Dominant | C/C | 140 (84.8) | 16 (66.7) | 1.00 | 0.042a | 143.7 |
| C/T‐T/T | 25 (15.2) | 8 (33.3) | 2.80 (1.08–7.24) | |||
OR, odds ratio; 95% CI, 95% Confidence Interval; AIC, Akaike Information Content; AF, Atrial fibrillation; CABG, coronary artery bypass grafting; CAD, Coronary Artery Disease.
Statistically significant at P < 0.05.
Discussion
In this study we demonstrated that the 4q25 risk allele rs10033464 (T) allele is significantly associated with increased risk of AF, more precisely with post‐CABG AF, whereas the rs2200733 (T) with lone AF, in our cohort of indo‐aryan ethnicity.
The rs2200733 and rs10033464 were first identified as AF risk conferring variants in a GWAS in the Icelandic population 5. Moreover, the haplotype identified by rs2200733 was found to confer a relatively higher risk (OR 1.72), whereas the haplotype identified by rs10033464 conferred a more modest risk (OR 1.39). The study also replicated the findings in the North Europeans as well as Chinese population in Hong Kong.
Another large‐scale replication 11 on four cohorts of European descent the FHS, the Rotterdam Study (RS), Vanderbilt AF registry, and the German AF cohort echoed the initial GWAS findings wherein the rs2200733 was associated with AF in all the four cohorts with an OR ranging from 1.37 to 2.52. There was a significant association between AF and rs10033464 in Framingham (OR 1.34; 95% CI 1.03–1.75; P = 0.031) and the German AF Network (OR 1.30; 95% CI 1.13–1.51; P = 0.0002), but not Vanderbilt AF registry (OR 1.16; 95% CI 0.86–1.56; P = 0.33) or the RS (OR, 1.17; CI 0.99–1.38; P = 0.07).
In this study the rs10033464 conferred a higher risk (OR, 2.59; AF vs. Controls: OR, 2.73; AF vs. CAD) than that of the GWAS study or the replication study in European cohorts; moreover, the risk was significantly higher for post‐CABG AF (OR, 4.73; post‐CABG AF vs. Controls: OR, 2.73; post‐CABG AF vs. CAD), whereas the rs2200733 only conferred a higher risk for lone AF (OR, 2.80 lone AF vs. CAD).
Our study is also in concordance with the findings of Kiliszek et al. 6 who found that rs10033464 was associated with increased risk of AF (OR 1.69, 95% CI 1.23–2.33, P = 0.00125), the strongest association was seen with the rs2200733 in the same study, which also favored increased number of episodes of AF/month. In the same study SNPs at 16q22 and 1q21 locus also had a significantly higher frequency in patients qualified for catheter ablation of AF than the control group.
Interestingly findings by Virani and colleagues 13 indicated rs10033464 to be significantly associated with PoAF after CABG (OR1.47, 95% CI 1.05–2.06, respectively) in the fully adjusted model. The use of beta blockers (BB) appeared to increase the risk of PoAF in the carriers of the minor allele for rs10033464: OR 1.89, 95% CI 1.22–2.93 for those receiving preoperative BB and OR 1.04, 95% CI 0.65–1.65 for those not receiving BB. Cox regression analyses revealed rs10033464 was also associated with an increased risk of long‐term AF development (HR‐ 1.28, 95% CI 1.00–1.66, respectively). Our study is in complete agreement with the above findings, wherein the rs10033464 showed significant association with post‐CABG AF.
In variance with the findings by Body et al. 10 which indicated the rs2200733 to predict for postoperative AF, in our study the rs2200733 did not associate with postoperative AF, but lone AF. Again rs10033464 which was significantly associated with post‐CABG AF in our study, however, did not associate with postoperative AF either in the discovery (OR 1.21, P = 0.28) or in the validation cohort (OR 1.41, P = 0.16) of the study by Body et al.
Our study is also at variance with the observations by Delaney et al. 14 in the African‐American (AA) population. In this study of the six previously reported SNPs associated with AF, only three were found to be associated with AA albeit with a reduced effect. The rs10033464 was not associated with AF in this population, the rs2200733 which was the index SNP in European American population showed less significant association as compared to rs4631108 which emerged as the index SNP. In our study the rs2200733 was significant (OR 2.73; P = 0.05) in individuals with lone AF, but not post‐CABG AF.
Lubitz et al. 15 also demonstrated the rs2200733 with increased AF risk (OR 1.80, 95% CI 1.50–2.15, P = 1.2 × 10−20) but not the rs10033464 (OR 1.07, 95%CI 0.84–1.35, P = 0.59), whereas our results indicate the opposite.
Comparison of the minor allele frequency for rs10033464 in this study population and that of other population in the Hap‐Map database indicates the allelic frequency in our study (0.12) to be closest to Utah residents with Northern and Western European ancestry (CEU‐ 0.10) and of Gujarati Indians in Houston (GIH‐ 0.18) on Hap‐Map data, whereas differences in allelic frequencies are seen among populations like Han Chinese in Beijing (HCB‐0.24), Yoruba in Ibadan (YRI‐0.24), and Japanese in Tokyo, Japan (JPT‐0.27).
Similar findings were also seen for the rs2200733 variant, where the allelic frequency in this study (0.13) is closest to CEU population (0.11) and GIH (0.14), but significantly different from JPT (0.47), HCB (0.42), and YRI (0.25).
The differences seen between the above‐mentioned studies and this study could be due to possible differences in the linkage disequilibrium (LD) and the haplotype blocks defined by these SNP could vary by ethnicity.
In conclusion, although this study has limitations, primarily in terms of sample size, nevertheless, this is the first study to determine the frequency and association of the 4q25 variants with AF risk in the Indian population. These variants could stratify postoperative, long‐term AF risk (PoAF) and also potentially affect the choice of preoperative therapy to decrease the risk of PoAF, and larger studies as well as fine mapping of the 4q25 locus to understand the LD between variants are warranted.
Author Contributions
AAB conceived the study, designed the experimental work, analyzed results, and wrote the manuscript. BRD conceived the study and reviewed the manuscript. TS and GJ participated in selection of patients and reviewed the manuscript. GJ and AN performed the experimental work. All authors have read and approved the final manuscript.
Funding Source
No funding for this project was received from any external source.
References
- 1. Fuster V, Ryden LE, Asinger RW, et al. ACC/AHA/ESC Guidelines for the Management of Patients With Atrial Fibrillation: Executive Summary A Report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines and the European Society of Cardiology Committee for Practice Guidelines and Policy Conferences (Committee to Develop Guidelines for the Management of Patients With Atrial Fibrillation) Developed in Collaboration With the North American Society of Pacing and Electrophysiology. Circulation 2001;104:2118–2150. [PubMed] [Google Scholar]
- 2. Wakili R, Voigt N, Kaab S, Dobrev D, Nattel S. Recent advances in the molecular pathophysiology of atrial fibrillation. JCI 2011;121:2955–2968. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Piccini JP, Hammill BG, Sinner MF, et al. Incidence and prevalence of atrial fibrillation and associated mortality among medicare beneficiaries, 1993–2007. Circ Cardiovascular Qual Outcomes 2012;5:85–93. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Benjamin EJ, Levy D, Vaziri SM, D'Agostino RB, Belanger AJ, Wolf PA. Independent risk factors for atrial fibrillation in a population‐based cohort: The Framingham Heart Study. JAMA 1994;271:840–844. [PubMed] [Google Scholar]
- 5. Gudbjartsson DF, Arnar DO, Helgadottir A, et al. Variants conferring risk of atrial fibrillation on chromosome 4q25. Nature 2007;448:353–357. [DOI] [PubMed] [Google Scholar]
- 6. Kiliszek M, Franaszczyk M, Kozluk E, et al. Association between variants on chromosome 4q25, 16q22 and 1q21 and atrial fibrillation in the Polish population. PLoS ONE 2011;6:e21790. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Husser D, Adams V, Piorkowski C, Hindricks G, Bollmann A. Chromosome 4q25 variants and atrial fibrillation recurrence after catheter ablation. JACC 2010;55:747–753. [DOI] [PubMed] [Google Scholar]
- 8. Viviani Anselmi C, Novelli V, Roncarati R, et al. Association of rs2200733 at 4q25 with atrial flutter/fibrillation diseases in an Italian population. Heart 2008;94:1394–1396. [DOI] [PubMed] [Google Scholar]
- 9. Shi L, Li C, Wang C, et al. Assessment of association of rs2200733 on chromosome 4q25 with atrial fibrillation and ischemic stroke in a Chinese Han population. Hum Genet 2009;126:843–849. [DOI] [PubMed] [Google Scholar]
- 10. Body SC, Collard CD, Shernan SK, et al. Variation in the 4q25 chromosomal locus predicts atrial fibrillation after coronary artery bypass graft surgery. Circulation Cardiovascular Genetics 2009;2:499–506. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Kaab S, Darbar D, van Noord C, et al. Large scale replication and meta‐analysis of variants on chromosome 4q25 associated with atrial fibrillation. Eur Heart J 2009;30:813–819. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Sole X, Guino E, Valls J, Iniesta R, Moreno V. SNPStats: A web tool for the analysis of association studies. Bioinformatics 2006;22:1928–1929. [DOI] [PubMed] [Google Scholar]
- 13. Virani SS, Brautbar A, Lee V‐V, et al. Usefulness of single nucleotide polymorphism in chromosome 4q25 to predict in‐hospital and long term development of atrial fibrillation and survival in patients undergoing coronary artery bypass grafting. Am J Cardiol 2011;107:1504–1509. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Delaney JT, Jeff JM, Brown NJ, et al. Characterization of genome‐wide association‐ identified variants for atrial fibrillation in African Americans. PLoS ONE 2012;7:e32338. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Lubitz SA, Sinner MF, Lunetta KL, et al. Independent susceptibility markers for atrial fibrillation on chromosome 4q25. Circulation 2010;122:976–984. [DOI] [PMC free article] [PubMed] [Google Scholar]